| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Academic Department of Radiology, Royal Marsden Hospital, Downs Rd., Sutton,
Surrey, United Kingdom SM2 5PT.
2 Department of Radiology, Hillingdon Hospital, Pield Health Rd., Middlesex,
United Kingdom UB8 3NN.
3 Department of Medical Statistics, Royal Marsden Hospital, Downs Rd., Sutton,
Surrey, United Kingdom SM3 6AL.
Received August 15, 2005;
accepted after revision February 1, 2006.
Address correspondence to D.-M. Koh.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Sixty-one patients with a potentially resectable hepatic metastasis of colorectal cancer consecutively underwent breath-hold T1-weighted MRI in the axial and coronal planes 30 minutes and 24 hours after administration of MnDPDP. For each lesion, the presence or absence of rim enhancement and segmental enhancement 30 minutes and 24 hours after contrast administration was recorded. These features were evaluated separately for lesions 10 mm in diameter or larger and lesions smaller than 10 mm. The nature of each lesion was determined at histopathologic examination (n = 29) and on follow-up imaging (n = 32).
RESULTS. Two hundred thirty lesions were identified at MRI: 210 lesions were metastatic, and 20 were benign. Rim enhancement was observed around 22 of 210 (10%) of the metastatic lesions at 30 minutes and 199 of 210 (95%) of metastatic lesions at 24 hours. Rim enhancement at 24 hours had 94.8% (95% CI, 91.8-97.8%) sensitivity, 90.0% (68.3-98.8%) specificity, 99.0% (97.6-100%) positive predictive value, 62.1% (42.3-79.3%) negative predictive value, and 94.3% (91.4-97.3%) diagnostic accuracy for metastasis. Segmental enhancement was infrequently seen (34/210; 16%) at 24 hours but had 100% (89.7-100%) positive predictive value for metastasis.
CONCLUSION. Rim and segmental enhancement at MRI 24 hours after MnDPDP administration enabled accurate characterization of hepatic colorectal metastasis. These features may aid in preoperative mapping of hepatic tumor burden and disease distribution in patients with colorectal cancer.
Keywords: abdominal imaging • liver • MR contrast agents • MRI • oncologic imaging
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Preoperative chemotherapy can result in substantial downstaging of hepatic metastasis. Oxaliplatin-based chemotherapy regimens have been shown to reduce tumor burden and can convert as many as one third of unresectable cases to resectability with curative intent [6-9]. Therefore, the ability to show both treatment response and the presence of measurable foci after treatment becomes important. However, detection and characterization of hepatic lesions measuring less than 1 cm in diameter remains challenging.
Mangafodipir trisodium (MnDPDP, Teslascan, Nycomed Amersham) is a hepatocytespecific MRI contrast agent with a mechanism of action different from that of gadolinium-based contrast agents. After IV administration of MnDPDP, maximal T1-weighted liver enhancement occurs in 15 minutes and persists for several hours [10]. At maximum liver enhancement, the contrast-to-noise ratio between nonhepatocellular lesions and liver parenchyma is fivefold greater than that on unenhanced images [11]. During this period, nonhepatocellular lesions, such as those of colorectal metastasis, appear as low-signalintensity lesions against enhanced liver parenchyma [12-19]. MnDPDP contrast material normally clears from hepatocytes 24 hours after administration, but delayed clearance has been reported in liver parenchyma surrounding metastatic lesions [16, 20].
Most of the published descriptions of MnDPDP MRI in the evaluation of metastasis were performed at maximum liver parenchymal enhancement (30 minutes-3 hours) [12, 14-17, 19]. The appearance of metastatic lesions 24 hours after clearance of MnDPDP from hepatocytes has been reported in only small series of patients [16, 20]. One of the reasons for the paucity of studies of MnDPDP MRI 24 hours after contrast administration may relate to the practicalities of performing two MRI examinations in a 24-hour period.
Colorectal metastatic lesions have been reported to show rim and segmental enhancement 24 hours after contrast administration [16, 20]. To our knowledge, however, in no single study have investigators fully described the MnDPDP MRI appearances of colorectal hepatic metastatic lesions at peak hepatic enhancement (30 minutes) and after contrast clearance from hepatocytes (24 hours). Furthermore, the frequency and diagnostic accuracy of rim and segmental enhancement are uncertain. The aim of this study was to determine the diagnostic accuracy of rim and segmental enhancement in MnDPDP MRI detection of colorectal hepatic metastasis.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Sixty-one patients (42 men and 19 women; mean age, 64.5 years; age range, 19-77 years) with potentially operable hepatic metastasis of colorectal cancer were consecutively enrolled over a 14-month period and were prospectively examined with MnDPDP-enhanced MRI. We believed this study population was likely to have histopathologic validation of the MRI observations.
The inclusion criteria were pathologic proof of colorectal cancer; contrast-enhanced MDCT visualization of hepatic lesions consistent with metastasis; and presence of metastatic lesions deemed potentially resectable with extended hepatectomy, hepatectomy, or segmentectomy, defined on CT as any number of hepatic metastatic lesions with sparing of at least two contiguous liver segments. The exclusion criteria were history of other malignant disease and the presence of widespread liver metastasis with sparing of fewer than two contiguous liver segments.
MRI
MR images were acquired with one of two 1.5-T MRI systems (Magnetom Vision,
Siemens Medical Solutions; Intera, Philips Medical Systems). Unenhanced axial
T2-weighted turbo spin-echo sequences (TR/first-echo TE, second-echo TE,
1,800/80 and 180; effective TR, 29; number of excitations, 1; flip angle,
90°; field of view, 400 mm; matrix size, 176 x 256; section
thickness, 7 mm; slice gap, 1 mm; acquisition time, 42 seconds) and
breath-hold axial and coronal T1-weighted gradient-echo sequences (128/4.6;
flip angle, 75°; field of view, 400 cm; matrix size, 256 x 256;
section thickness, 7 mm; slice gap, 1 mm; acquisition time, 20-56 seconds)
were performed before contrast administration.
After acquisition of the unenhanced images, the patient was taken out of the MR unit to receive MnDPDP (5 mmol/kg body weight) by slow IV infusion over 20 minutes. Breath-hold axial and coronal T1-weighted gradient-echo sequences (128/4.6; flip angle, 75°; field of view, 400 cm; matrix size, 256 x 256; section thickness, 7 mm; slice gap, 1 mm; acquisition time, 20-56 seconds) were performed within 30 minutes of completion of contrast infusion during maximum hepatic parenchymal enhancement. The patients returned 24 hours (range, 22-26 hours) after contrast administration for further imaging after washout of contrast material from the hepatic parenchyma. The T1-weighted imaging protocol was the same as that used for acquisition of the first set of contrast-enhanced images. The MRI examination time before contrast administration was typically 15 minutes. The MRI examination time after contrast administration was typically 10 minutes.
Image Evaluation
MnDPDP MR images were reviewed by three radiologists in consensus. Two of
these radiologist had more than 10 years of experience in body MRI, and the
other radiologist had 5 years of experience in body MRI. The reviewers were
blinded to all other imaging and histopathologic findings. For each patient,
the following were recorded: number of lesions identified; size of each lesion
measured on the workstation (largest transverse diameter to the nearest
millimeter) and whether the lesion was 10 mm or greater or less than 10 mm in
diameter; signal intensity of each lesion compared with the hepatic
parenchyma; border appearance of each lesion recorded as smooth, blurred,
irregular, or irregular with satellitism (Fig.
1A,
1B,
1C,
1D) on 30-minute images when
liver parenchymal enhancement was maximum, allowing clear definition of the
margins of the lesions; absence or presence of rim enhancement and segmental
enhancement at 30 minutes; and absence or presence of rim enhancement and
segmental enhancement at 24 hours (Figs.
2 and
3A,
3B,
3C).
|
|
|
|
|
|
|
|
At the end of chemotherapy, CT and MnDPDP MRI were repeated for all patients. Twenty-nine patients underwent surgery, and results of examination of the surgical specimens provided histologic validation. Surgery was performed 4-6 weeks after completion of chemotherapy.
The other 32 patients did not undergo surgery for one or more of the following reasons: MRI showed more extensive disease than initially suspected, contraindicating surgery; MRI showed disease progression after chemotherapy, contraindicating surgery; or the clinical status of the patient deteriorated, precluding surgery. These patients underwent follow-up with imaging.
Serial follow-up imaging was performed every 3-6 months with a combination of CT, MnDPDP MRI, and gadolinium-enhanced MRI. A lesion was defined as malignant if the maximum transverse diameter had increased at least 20% in the follow-up period (median, 14.2 months; range, 12-18 months). Benign lesions (cyst, hemangioma, focal nodular hyperplasia) were defined by stability in size and typical appearance on contrast-enhanced CT scans and unenhanced T1- and T2-weighted MR images.
Statistical Analysis
Statistical analysis was performed with SPSS software (version 11.5, SPSS).
The sensitivity, specificity, positive predictive value (PPV), negative
predictive value (NPV), and diagnostic accuracy of the enhancement patterns
for characterizing metastasis were computed and reported with 95% CIs.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Validation of lesions according to findings at histopathologic examination and follow-up imaging is summarized in Figure 2. Two hundred ten (91%) of the 230 lesions were metastatic; 165 of these lesions were 10 mm in diameter or larger, and 45 were smaller than 10 mm in diameter.
In the 29 patients who underwent hepatic surgery, 53 metastatic and 13 benign lesions (12 cysts and one hemangioma) were identified at histopathologic examination after surgery, corresponding to lesions identified on imaging. No lesion was found at surgery that was not detected at MRI. The presence of two lesions was established at percutaneous biopsy. These lesions were found in two patients who had a long interval (> 5 years) between management of the primary tumor and appearance of liver metastasis.
In the 32 patients who did not undergo surgery or biopsy, 155 metastatic and seven benign lesions (six cysts and one case of focal nodular hyperplasia) were recognized through comparison with follow-up images.
Number and Size of Lesions
The mean number of lesions found in each patient was 3.8, ranging from one
to 12. Overall, mean lesion size was 25.4 mm (range, 1-97 mm). The mean size
of metastatic lesions was 35.3 mm, and the mean size of benign lesions was
22.6 mm. Of the metastatic lesions, 165 of 210 (79%) were 10 mm in diameter or
larger, and 45 of 210 (21%) were smaller than 10 mm. Of the 20 benign lesions,
17 of 20 (85%) were 10 mm in diameter or larger, and three of 20 (15%) were
smaller than 10 mm.
Signal Intensity and Border Appearance
Metastatic lesions visible on unenhanced images appeared hypointense on
T1-weighted images and of variable hyperintensity on T2-weighted images. All
metastatic lesions were hypointense on T1-weighted images immediately after
administration of MnDPDP contrast material. Ninety percent (190/210) of
metastatic lesions had irregular borders or irregular borders with
satellitism. Among metastatic lesions 10 mm in diameter or larger, 88%
(145/165) had irregular borders, and 8% (14/165) had irregular borders with
satellitism. In comparison, among metastatic lesions smaller than 10 mm in
diameter, 69% (31/45) of the lesions had irregular borders, and 29% (13/45)
had blurred margins.
Enhancement with MnDPDP
The frequency of rim and segmental enhancement on MRI 30 minutes and 24
hours after MnDPDP contrast administration is summarized in Tables
1 and
2. Only 22 of 210 (10%)
metastatic lesions exhibited rim enhancement 30 minutes after MnDPDP contrast
administration. Segmental enhancement was not observed at 30 minutes. None of
the benign lesions exhibited rim enhancement. Twenty-four hours after MnDPDP
contrast administration, 199 of 210 (95%) metastatic lesions exhibited rim
enhancement. All 22 of 210 metastatic lesions that exhibited rim enhancement
at 30 minutes also exhibited rim enhancement at 24 hours (Fig.
3A,
3B,
3C). Of these, 43 of 45 (96%)
were smaller than 10 mm in diameter (Fig.
4A,
4B,
4C). Rim enhancement was well
seen on axial images in most cases (183/199; 92%) but was best appreciated on
coronal images in a small proportion (16/199; 8%) of cases.
|
|
|
|
|
Segmental enhancement was found in only 34 of 210 (16%) cases. In four large metastatic lesions, segmental enhancement made it difficult to discern the presence of rim enhancement (Fig. 5A, 5B). Segmental enhancement was the only detectable feature in one patient with a small histologically proven metastatic lesion within the right lobe of the liver (Fig. 6A, 6B, 6C).
|
|
|
|
|
|
Among the benign lesions, 17 of the 18 cysts did not exhibit rim or segmental enhancement at 24 hours. A 2-cm cyst lying below the diaphragm in the right lobe was recorded as exhibiting rim enhancement on coronal images. However, findings at repeated review of the images showed that the perception of rim enhancement might have been caused by a thin layer of hepatic parenchyma lying between the low signal intensity of the cyst on one side and air in the lung base on the other. We nevertheless decided that a false-positive finding of rim enhancement had occurred. A hemangioma exhibited incomplete rim enhancement at 24 hours. The focal nodular hyperplasia appeared isointense in relation to the surrounding liver immediately after contrast administration and remained uniformly hyperintense on T1-weighted imaging at 24 hours.
Diagnostic Accuracy and Predictive Value of
Enhancement Patterns Overall, rim enhancement at 24 hours had
94.8% (95% CI, 91.8-97.8%) sensitivity, 90.0% (68.3-98.8%) specificity, 99.0%
(97.6-100%) PPV, 62.1% (42.3-79.3%) NPV, and 94.3% (91.4-97.3%) diagnostic
accuracy for metastasis. For lesions 10 mm or more in diameter, rim
enhancement on 24-hour imaging had 94.6% (91.1-98.0%) sensitivity, 88.2%
(63.6-98.5%) specificity, 98.7% (97.0-100%) PPV, 62.5% (40.6-81.2%) NPV, and
94.0% (90.5-97.4%) diagnostic accuracy for metastasis. For lesions smaller
than 10 mm in diameter, rim enhancement at 24 hours had 95.6% (84.9-99.5%)
sensitivity, 100% (29.3-100%) specificity, 100% (91.8-100%) PPV, 60%
(5.3-85.4%) NPV, and 95.8% (85.8-99.5%) diagnostic accuracy for
metastasis.
Segmental enhancement was infrequently found on 24-hour imaging. Overall, segmental enhancement had 16.2% (11.2-21.2%) sensitivity, 100% (83.2-100%) specificity, 100% (89.7-100%) PPV, 10.2% (6.0-14.4%) NPV, and 23.5% (18.0-29.0%) accuracy. The PPV for metastatic lesions 10 mm in diameter or larger was 100% (87.7-100%). The PPV for metastatic lesions smaller than 10 mm in diameter was 100% (54.1-100%).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Previous studies [16, 17] have shown that MRI performed after MnDPDP contrast administration can be used to differentiate lesions of hepatocellular from those of nonhepatocellular origin on the basis of the presence or absence of uptake of contrast medium. Hepatocellular lesions such as focal nodular hyperplasia, adenoma, regenerative nodules, and hepatocellular carcinoma exhibit MnDPDP contrast uptake and T1 signal enhancement. Metastasis arising from neuroendocrine tumors intriguingly also exhibits contrast uptake by an unknown mechanism [22]. In contrast, nonhepatocellular lesions such as metastatic lesions of nonneuroendocrine orgin, abscess, and hemangioma do not enhance [15-17].
Hepatocyte-selective gadolinium chelates such as gadobenate dimeglumine (Multi-Hance, Bracco) and ethoxybenzyl gadopentetate dimeglumine (Primovist, Schering) are administered as an IV bolus and behave as lowmolecular-weight extracellular contrast agents immediately after injection, enabling dynamic imaging. However, a hepatospecific phase occurs approximately 30 minutes after injection because of percentage biliary excretion, which may further aid lesion detection and characterization. Despite the theoretic advantages of the newly approved hepatocyte-selective gadolinium chelates, to our knowledge no typical features of hepatic metastasis in the liver-specific phase of gadobenate dimeglumine or ethoxybenzyl gadopentetate dimeglumine imaging have been described. Imaging performed 24 hours after administration of these contrast agents also has not been reported.
In patients with colorectal hepatic metastasis, knowledge of the distribution and extent of disease is important for treatment planning. Dynamic gadopentetate dimeglumine-enhanced MRI is widely acknowledged to be the investigation of choice for the characterization of focal liver lesions. Diagnostic uncertainty, however, particularly for small lesions, can result from breathing or movement artifacts during dynamic imaging performed with nonspecific gadolinium chelates and hepatocyte-selective gadolinium chelates. Despite the use of MDCT [23], many lesions measuring less than 1 cm in diameter remain too small to characterize. In one study [24], 14% of lowdensity lesions smaller than 1 cm in diameter on CT of patients with colorectal cancer were found to be metastatic at follow-up imaging. Thus, detection and characterization of hepatic lesions smaller than 1 cm in diameter can be challenging.
Thirty minutes after MnDPDP administration, strong hepatic parenchymal enhancement allows clear definition of nonhepatocellular lesions. Not surprisingly, previous studies [12-14] have shown that MnDPDP imaging at maximum liver enhancement was comparable with or superior to contrast-enhanced CT in the detection of colorectal hepatic metastasis. One study [19] also showed that MnDPDP MRI was more sensitive than PET in the detection of hepatic metastatic lesions smaller than 10 mm in diameter.
Our study confirmed that all colorectal metastatic lesions appear hypointense on T1-weighted MR images after MnDPDP contrast administration. Metastatic lesions also frequently had irregular borders or irregular borders with satellitism on imaging 30 minutes after contrast administration. Border characterization was less certain for metastatic lesions smaller than 10 mm in diameter because a large proportion exhibited blurred outlines related to partial volume averaging. Nevertheless, an irregular outline was identifiable in 70% of small metastatic lesions.
In our study, rim enhancement was seen in only 10% of metastases 30 minutes after MnDPDP administration. This finding is not surprising because the brightly enhancing hepatic parenchyma obscured visualization of rim enhancement. This finding is in accordance with the low incidence of rim enhancement found in studies [12-14] in which only early imaging (10 minutes-2 hours after MnDPDP administration) was performed. Hence, rim enhancement 30 minutes after contrast administration is not a reliable feature for discriminating metastasis from other lesions.
In comparison with early imaging at 30 minutes, imaging 24 hours after contrast administration showed rim enhancement in 95% of large metastatic lesions and in 96% of small metastatic lesions. Rim enhancement is believed to result from local biliary stasis due to compression of adjacent hepatic parenchyma by metastasis [17] and hence would be better shown after washout of contrast medium from the hepatic parenchyma. We found rim enhancement at 24 hours to have a high sensitivity, PPV, and diagnostic accuracy for metastasis, including metastatic lesions smaller than 10 mm in diameter.
Segmental enhancement was a feature found only on imaging 24 hours after contrast administration. This appearance might have been caused by compression of a segmental or subsegmental biliary duct by metastasis, the result being delayed clearance of contrast medium from the liver segment. We observed that segmental enhancement increased the conspicuity of small metastatic lesions by enlarging the visible abnormality, thus enhancing detection (Fig. 6A, 6B, 6C). For one lesion, segmental enhancement was the only imaging clue to underlying metastasis. However, segmental enhancement was seen in only 16% of metastases in our study. The low sensitivity may not reflect failure of the diagnostic test because not all metastatic lesions are expected to lie adjacent to bile ducts. The high specificity has to be interpreted with caution because of the small number of benign lesions in our study group. Nevertheless, segmental enhancement was found to have a very high PPV for metastasis.
On the basis of our findings, we recommend the use of MnDPDP-enhanced MRI for detection and characterization of hepatic metastasis in patients with colorectal cancer. Early imaging 30 minutes after contrast administration allows clear definition of the outline of metastatic lesions, facilitating lesion detection [12-14]. Rim and segmental enhancement on delayed imaging 24 hours after contrast administration can be used for lesion characterization.
There were a few limitations to the current study. First, we established two MnDPDP MRI enhancement features that appear helpful for characterization of colorectal hepatic metastasis. However, we did not test the hypothesis in the general population, and further studies may be required. Second, the number of benign lesions was relatively small in comparison with the number of metastatic lesions in our study group. However, the numbers reflected the study population, in which the inclusion criteria were based on identifying patients with metastasis who were expected to benefit from surgical resection. Nevertheless, rim and segmental enhancement on 24-hour imaging were found to have diagnostic utility in the detection of metastasis.
Third, three reviewers were involved in consensus scoring of the images, but interobserver agreement was not tested. This factor can be assessed in future studies. Fourth, lesions were classified as benign on follow-up imaging on the basis of stability in size. Patients in the follow-up group usually had a clear increase in size of metastatic lesions after chemotherapy or had more extensive disease, both of which were contraindications to surgical resection. In the relatively long serial follow-up imaging period, it is unlikely that such a malignant lesion would remain static in size, although there was a small possibility that this phenomenon could have occurred. Finally, we were unable to obtain 100% histopathologic comparison for all lesions detected because only 30% of all metastatic lesions detected on MRI were surgically resected. Follow-up imaging findings were available, however, for all patients and provided indirect evidence of the presence or absence of malignant disease.
In conclusion, colorectal hepatic metastasis had characteristic features at MnDPDP MRI 24 hours after contrast administration. Rim enhancement was frequently found and had a high sensitivity, PPV, and diagnostic accuracy for metastasis. Segmental enhancement, although infrequently found, had a high PPV.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
2 cm) hepatic lesions in
colorectal cancer patients: detection and characterization on mangafodipir
trisodium-enhanced MRI. AJR 2004;182
: 1233-1240 1.5 cm) liver metastases:
is thinner collimation better? Radiology2002; 225:137
-142
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
